A modeling framework for quantifying spatial recruitment dynamics using abundance estimation and sibship analysis: code and simulation study output
Data files
Aug 26, 2024 version files 2.66 MB
-
README.md
-
SimResults_MultiTimeStep.csv
-
SimResults_singleCohort.csv
Abstract
Quantifying recruitment at the sibling group offers a powerful methodology for understanding density-dependent and environmental drivers of recruitment. We propose a modeling framework that combines sibship and abundance estimation datasets to estimate mean sibling group size, sibling group size process error, environmental and density-dependent effects on sibling group size, dispersal, and mortality rate. Geographic states in the model consist of discrete habitat patches connected via dispersal. Simulations were used to investigate the influence of sampling processes and sibling group size on parameter estimation within our modeling framework. Mean sibling-group size, environmental effects on recruitment, and dispersal rate among habitat patches were estimated with high accuracy under a wide range of sampling conditions, including imprecise out-of-model estimates of capture probability and subsampling both within and among habitat patches. Density-dependent effects on recruitment and process error tended to be estimated with lower accuracy, though accuracy improved as sibling group size or sampling intensity increased. The main contribution of this research is a flexible quantitative modeling framework for parameterizing mechanistic models of recruitment dynamics with empirical sibship data.
README: A Modeling Framework for Quantifying Spatial Recruitment Dynamics Using Abundance Estimation and Sibship Analysis: Code and Simulation Study Output
https://doi.org/10.5061/dryad.2fqz612zd
Description of the data and file structure
The simulation results provided summarizes simulation output obtained using the associated software code provided (R scripts and Stan model code). The raw simulation output (stan model for each model run) was summarized by 1) extracting the parameter values and model diagnostics of interest from each stan model fit to a simulated dataset and 2) calculating the relative error and precision for each simulation.
Files and variables
File: SimResults_MultiTimeStep.csv
Description:
Variables
- mean: mean of the posterior distribution
- 10%: 10th quantile of the posterior distribution
- 90%: 90th quantile of the posterior distribution
- TrueSimValue: value used in the data-generating simulation model
- posterior_sd: sd of the posterior distribution
- CICoverage: does the TrueSimValue fall between 10% and 90% (1=Yes; 0=No)
- CIPrecision: precision of the parameter estimate
- RMSE: root-mean-square-error of the parameter estimate
- Bias: relative error of the parameter estimate
- maxRhat: maximum r-hat from the stan model output
- min_n_eff: minimum effective sample size from the stan model output
- passchecks: did the model pass convergence checks? (1=Yes, 0=No)
- parm: parameter name
- simRep: simulation rep ID
- nSamples: number of samples per year
- p_sd: sample-to-sample variance in detection probability
- npatch: number of habitat patches in the simulated metapopulation
- nfg: number of family groups per year
- a: recruitment rate
- a_dense: density-dependent effects on recruitment rate
- a_sd: family group to family group variability in recruitment rate
- q: dispersal rate
- propPatchSurveyed: proportion of each sampled patch that was surveyed
- a_beta: environmental effects on recruitment rate
- captureProb_int: mean capture probability
- nFG_obs: number of family groups detected
- avg_n_per_fg: average number of captured individuals per family group
- total_catch_obs: total number of captured individuals
File: SimResults_singleCohort.csv
Description:
Variables
- mean: mean of the posterior distribution
- 10%: 10th quantile of the posterior distribution
- 90%: 90th quantile of the posterior distribution
- TrueSimValue: value used in the data-generating simulation model
- posterior_sd: sd of the posterior distribution
- CICoverage: does the TrueSimValue fall between 10% and 90% (1=Yes; 0=No)
- CIPrecision: precision of the parameter estimate
- RMSE: root-mean-square-error of the parameter estimate
- Bias: relative error of the parameter estimate
- maxRhat: maximum r-hat from the stan model output
- min_n_eff: minimum effective sample size from the stan model output
- passchecks: did the model pass convergence checks? (1=Yes, 0=No)
- parm: parameter name
- simRep: simulation rep ID
- nSamples: number of samples per year
- p_sd: sample-to-sample variance in detection probability
- npatch: number of habitat patches in the simulated metapopulation
- nfg: number of family groups per year
- a: recruitment rate
- a_dense: density-dependent effects on recruitment rate
- a_sd: family group to family group variability in recruitment rate
- q: dispersal rate
- propPatchSurveyed: proportion of each sampled patch that was surveyed
- a_beta: environmental effects on recruitment rate
- captureProb_int: mean capture probability
- nFG_obs: number of family groups detected
- avg_n_per_fg: average number of captured individuals per family group
- total_catch_obs: total number of captured individuals
Code/software
R and the packages listed in the code are required to run the scripts.
SimulationSummaryandPlots_MultiTimeStep_CJFAS_R07042024.R: summarize and plot simulation results from Lewandoski and Brenden (2024)
SimulationSummaryandPlots_SingleCohortCJFAS.R: summarize and plot simulation results from Lewandoski and Brenden (2024)
sim_singleCohort.R: simulate datasets and estimate parameters using Stan (single year run for each cohort)
sim_multiTimeStep.R: simulate datasets and estimate parameters using Stan (multi-year run for each cohort)
multitimestep.stan: Stan model for multi-time step model runs
singlecohort.stan: Stan model for single year model runs
SpatialPopDySimFunction_vDissertationChapter2.R: simulation function (called by sim_multiTimeStep.R)
SpatialPopDySimFunction_vDissertationChapter2_v_oneYr.R: simulation function (called by sim_singleCohort.R)
Methods
The simulation results were obtained using the code provided in the linked software related work (R code and Stan code provided).